| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.3310277319240384, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0176, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.5539650140867, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0223, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.416279485696912, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.0131, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.92211927330892, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0209, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.773399040125687, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.002, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.6503932165384456, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9539, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.219143676681775, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.888, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.9074303866806384, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9348, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 5.084742780533626, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9942, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.922188609323376, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9536, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 4.6161360497572765, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.9376, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.836280657898205, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8455, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 3.3475765151796835, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8465, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.812006112202477, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.871, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.301052036398465, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.7946, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.208387346864856, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.7332, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.336697083591116, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.7765, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.253031685193553, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7972, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 28.89751998836799, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7454, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1219.399671087463, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7813, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.8687972632165963, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.7354, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.907137607930398, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.7035, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 2.3211996211953796, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6445, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.4267741774585478, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.6627, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.1355836344893646, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6564, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.4324686987807933, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.7114, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.3054627994957466, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6458, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.1665825107556536, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6612, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.0653646732142628, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6389, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9426549245107123, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6027, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.0031299756100804, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.5574, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.8577295917991478, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.5607, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.8754224991833897, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5066, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.0364373817150851, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.5341, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.9757414507093809, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.5607, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.9148020456427475, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.4873, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 1.0030325891717722, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.564, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.9685020902194278, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.5296, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.888995210476553, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5211, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.072848290556435, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.4905, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 1.0496324671705435, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.4621, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.9844760999694531, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.423, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8041562724842711, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.3786, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.068075249758729, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.3874, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 1.211344200671008, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.3676, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 1.109593127362721, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.3807, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.9835519944720997, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.3675, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.946904254770148, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.3874, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.8503544094847885, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.3662, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.8718477101785147, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.356, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.8575772740880494, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.2892, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.7243817513291823, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.27, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.9759610361401863, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.2993, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.8964218318570011, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.2622, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.9470904012251518, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.2782, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.8263500861466068, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.2697, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.83699990653058, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.2073, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.7970990938820094, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.2679, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.8459005572678202, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.3047, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.7875190306363022, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.304, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.7637468746535009, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.185, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.70064023608572, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.1992, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.9042811360863167, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.1987, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 1.025824949915053, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.201, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.7641507051961434, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.1867, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.7497740208967364, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.1845, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.779800556072509, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.191, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.7931259235573362, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.1968, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.7558839152502762, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.1942, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.8127076740394721, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.1742, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.6967475558810442, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.1362, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.6127322796539928, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.1542, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.7174164591482776, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.1288, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.7718416906941996, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.116, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.7893737826947238, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.1276, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.6196087152921802, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.1416, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.6721041531860343, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.1519, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.7164936855269799, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.1222, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6821396600441599, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.1207, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6917119025509734, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.0908, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.5796453363283712, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.0937, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.6525959265094305, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.1052, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.5992297453671478, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.096, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.6221643996027728, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.1011, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.7346532024031716, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.083, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.6681341872972646, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.0947, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6496071515712458, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.0658, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.6131827491418356, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.0913, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.5364444549494034, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.0589, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5871805208175205, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.1016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.6376546058334409, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0795, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.5363126232439626, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.0701, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.47475857826005874, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.0636, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.46408013432183753, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.0651, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.4942073562494605, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.0675, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.5167214076500193, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.0569, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.5528173964684273, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.0593, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.6195909228875404, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.0547, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.44379736597934477, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.0499, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.534994585629485, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.0588, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.3899976705991816, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.0399, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.47749317716732687, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.0642, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.43298607031497116, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.0456, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.4752881903342076, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0658, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.3887530705398169, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.0508, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.42285053771235476, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.0628, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.3727130095326469, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.0361, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.3359928923843793, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.0385, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.3401401688909431, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.0399, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.3862502647717241, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.044, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.3785203408674366, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.0557, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.3748924116298517, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.0512, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.3134439887512741, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.0393, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.3060023540734267, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.0319, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.4026452889931919, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.0522, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.28136666505941504, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.0321, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.28941383864914, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.04, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.37953724203603134, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.0655, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.2925600622434635, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.0325, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.27008576143910745, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.0221, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.3373027438476517, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.0422, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.25613365174236163, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.0351, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.2815352624324449, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.0331, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.294124753558518, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.0389, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.30532136199040155, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.0368, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.3201687067241934, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.0467, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.27979442369625435, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.0349, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.3103650980354255, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.0346, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.29453151639746755, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.0421, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.3008195117667159, | |
| "learning_rate": 0.0, | |
| "loss": 0.0474, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 4.021832649172582e+16, | |
| "train_loss": 0.3273255914163131, | |
| "train_runtime": 1617.097, | |
| "train_samples_per_second": 2.54, | |
| "train_steps_per_second": 0.08 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.021832649172582e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |